Simulation As a Method To Support Complex Organizational
Transformations in Healthcare
Journal: 16th Americas Conference on Information Systems Manuscript ID: AMCIS-1037-2010.R1
Submission Type: Paper
Mini-Track: Role of technology in improving health care delivery processes < IT in Health Care (SIGHealth)
Simulation As a Method To Support Complex
Organizational Transformations in Healthcare
Diederik Rothengatter
University of Twente
d.c.f.rothengatter@utwente.nl
Christiaan Katsma
University of Twente
c.p.katsma@utwente.nl
Jos van Hillegersberg
University of Twente
j.vanhillegersberg@utwente.nl
ABSTRACT
In this paper we study the application of simulation as a method to support information system and process design in complex organizational transitions. We apply a combined use of a collaborative workshop approach with the use of a detailed and accurate graphical simulation model in a hospital that is in a major transition. The simulation represents the future situation of the hospital and enables the healthcare professionals to analyze and reflect on processes, planning, staffing and collaboration, and the new organizational processes and systems. We follow an action research setup in which, conducting iterative workshops, our simulation model is enhanced and fine-tuned. The data collected from the workshops reveal that the benefits of using simulations in complex organizational transitions include a substantial increase of awareness, insights and acceptance by the medical and administrative staff.
Keywords
Simulation, Organizational Change, Process Change, Healthcare, Hospital, Outpatient clinic, Information system, Reorganization, Participative design, Iterative analysis
INTRODUCTION
In this paper we study the application of simulation as a method to support information system -and process design in complex organizational transformations. We show the benefits over the mere analytic solutions by explicating this way of working and applying a software simulation to support problem analysis, and reframing the original development issues. The research has been carried out in the complex environment of outpatient clinics in a medium sized hospital in the Netherlands. In this paper we present a case study conducted from fall 2008 to winter 2009. In this case study the hospital is planning and implementing various changes in the building facilities, the information systems in place, the organizational structure, and the operational care and administrative processes.
Because of these substantial changes in delivering health care service to its patients, the hospital was faced with major uncertainties. Questions like ‘will we be able to deliver a continuous and stable output in patient treatments?’, ‘does the future building layout provide enough capacity and potential to expand production?’, ‘when do we need to register a patient, or will a patient register him/herself?’, and ‘who is responsible and accountable for things like patient throughput, and the utilization rate of shared resources?
The objective of this research is to show the added value of simulation as a method to analyze and reframe the original problem setting, into a constructive setting, where not only attention is paid to the problems, but also to opportunities to improve the processes, systems, and organization to ultimately improving the service delivery. We test this simulation method specifically in the health care domain because of the complex environment, the dynamic structure, and the high workload on the organization and staff.
RESEARCH APPROACH
The three authors have been involved in the case during the last 15 months and the type of research is best described as action research. In this period we have played both participant observer and expert roles. We have created different versions of a simulation model, that were used as input for design workshops in which:
• The future systems, processes/workflow, and organization structure were designed, • The simulation outcomes of prior design session were analyzed
• The impact on the organization’s systems, processes, and organization structure were evaluated.
In order to get the right and accurate input data for our software simulation model, we have interviewed numerous medical specialists, care professionals and administrative (management and support) staff during the case study. We used unstructured qualitative interviews and focused on the effects of the simulation, but also the insights of the interviewees. We also kept a diary to record observations, statements and make small notes on incidents or substantial events during the timeline or especially during the workshops. Every workshop was recorded and analyzed afterwards.
As part of the research project we developed a software simulation that analyzes physical patient movements. We translated key performance indicators in this health care setting to operational measures. These operational measures could be analyzed using the simulation model in the physical patient domain. Any suggestion for improvement or change initiative proposed by healthcare professionals was translated to the simulation model. By measuring the effectiveness and efficiency in the pre and post setting, the likeliness of success could be assessed. Furthermore this approach provided a collaborative setting, where the input from every process participant could be applied, analyzed, and evaluated. This resulted in a constructive, collaborative setting, where each participant was involved in the design of the future setup.
Simulation is extensively discussed in literature. In the Healthcare domain simulation has been extensively applied from the operations research perspective (Hans et al. 2008; van Oostrum et al. 2008). Mostly these simulations are detailed, based on factual or generated data, and supported by complex mathematical models. Practitioners in the health care domain state they support and believe the findings, but also express they have difficulties to adhere to and to really deploy the findings. In the Organizational Development domain several applications of interventions like gaming have shown the possibilities to contribute to these latter problems. (Boonstra 2004; French et al. 1984).
In this paper we especially chose a gaming and simulation approach in between the above described extremes. The simulation was not a mere change tool, but also not an expert mathematical model.
The simulation contained really substantial health care and process data, but also meaningful for the health care professionals. The way of working was a participative redesign set up that enabled practitioners to gradually understand and reflect on the data. Our assumption was this could work in two ways. The simulation becomes more realistic and the practitioners increase their knowledge, understanding and involvement. We intended to test this deployment of simulation already described in other domains outside the healthcare domain (Duke et al. 1981; Feinstein et al. 2002; Gredler 1992). Our research approach best can be described as an action research set-up in which the three authors were involved in the case for over 13 months. The healthcare professionals, via several participative workshops and sessions, were involved in analysis, design and game play of the simulation.
STRUCTURE
This paper is structured as follows:
Section 2 presents the case of the hospital, its background and in a chronological order the main observations of events, workshops and incidents Section 3 explains the problem of planning and coordination in the specific domain of hospitals and derive a model/foundation from literature. We use this model in the paper as a reference to put into perspective our case results. Section 4 explicates our analysis of the case results in the perspective of the model derived in section 3. We will focus on the viability of the simulation method and the effect and contributions during the process. Section 5 concludes our paper with a discussion and further recommendations.
CASE DESCRIPTION
This research has been carried out in a medium sized hospital (250 beds) in the Netherlands. The research project focused on the outpatient clinics serving over 210,000 patient treatments yearly. This hospital is facing many changes, caused by the future transfer to a brand new facility. In this new facility capacity in terms of available space is limited. The limitation of space is the reason that each medical specialism has to rethink their utilization of available rooms, waiting areas, and the use of shared resources. These items, in their turn, lead to rethinking the workflows, patient logistics, and scheduling. To induce the transfer from the current to the future situation, numerous projects are initiated. One project in specific is the patient logistics project for outpatient treatments. In this project the prosper allocation of physical space for patients spending their time awaiting their treatment is designed. Because of limitations on available square meters, waiting areas have to be shared amongst multiple departments. To account for an overflow of this capacity, a central waiting area is situated near the entrance of the hospital. This design is very similar to airport layouts. As will be discussed below, the future layout will have an enormous impact on the staff’s information systems, processes, and governance structure. They will have to rethink, and redesign the very basics of appointment scheduling, last-minute scheduling, and communication with other departments. At first glance the redesign op the scheduling and logistics process does not seem to impact the organization on a very high level, but in the next paragraph we will describe the major impact scheduling has on health care organizations as complex organizations.
COMPLEXITY OF HEALTH CARE ORGANIZATIONS
Of all highly complex service delivering organizations, hospitals can be ranked as one of the most complex (Hafferty et al. 2010). First, there is the difficult governance structure, in which healthcare managers and the board of directors have limited influence on the medical operations, and doctors and nurses only marginally have interest in the financial administration and the financial wellbeing of the hospital (Glouberman et al. 2001). These governance structure is setup to protect the doctor from behaving as an economic man (Harris 1977). This results in a fundamental setting of conflicting goals (Cutler 2002). Management and board try to optimize on output performance measures, doctors and nurses aim to improve the in-process performance measures. The aim of management could result in an overemphasis on KPI’s (Propper et al. 2008), the aim of the medical staff could result in a ‘flat of the curve medicine’ (Enthoven 1980). Second, there is the complexity of the medical process itself. Medical processes and diagnostics in particular, are characterized by their high stochasticity, variability, heterogeneity (Vandaele et al. 2003), complexity (Harper et al. 2002), and their lack of resources. Because of deviation caused by the medical process characteristics – let alone human influences – a high efficiency level is not within reach. In fact, the most cited problem in health care evidently is the lack of efficiency, or at least the need for improved efficiency, enforced by, among other things, government regulations. Taken from the health care organization viewpoint health care inefficiency can be measured by: (increase of) patient waiting time, (lack of) patient health improvement, or (low) percentage of resource utilization (Stiglic et al. 2005). Traditional approached to deal with these process characteristics are batching, buffering, and sequencing the patient treatments. Thirdly the hospital organization setup itself is a cause of inefficiency and complexity. Hospitals can be considered to be job shop organizations, with departmental budgeting, planning, and control. Significant resources are required to communicate planned and last-minute activities. Operating in this complex environment is tough, and the result expressed as efficiency is hardly ever satisfying (Cutler 2002; Feldstein 1974; Newhouse 1993). Key elements in operational efficiency are the means and success of coordination and communication (Malone et al. 1988). In order to streamline and improve the coordination and communication, the availability and use of information systems, the process efficiency, and the proper implementation of a suitable governance structure are essential. Considering these complex characteristics of health care organizations, it becomes clear why the described case setting will have an enormous impact on the processes. The unpredictability of the process makes it very hard for the medical staff to maintain a high level of efficiency. Current information systems, processes, and governance structures are the result of an evolutionary-like process. Most of them are tailored fit to the current operation. It has taken several years, sometimes even decades to get to this stage. Furthermore, proper scheduling using the current plan resources is the only method available to align the numerous actors involved in the treatment of a patient.
An example would be in place here. Scheduling, registering, and checking in a patient is one very common and basic process present in every hospital. Though, in a high level of abstraction the process is very generic, on the level of implementation, there are as many approaches to this process sequence as there are hospitals. The evolutionary design is the reason for the many variations, in the hospital case under review too. The new information system installed in the new hospital building will facilitate several new options for each separate activity (scheduling, registering, and checking in). And because of the emphasis on efficiency, and quality of service by the hospital management, ‘things’ need to be changed. One option available
is that patients schedule their meetings with physicians themselves. They can register their information in the system too, and check-in when they arrive at the hospital. An information system connected to the internet, and communicating with distributed terminals at several locations could facilitate this. The only problem then is that physicians don’t have detailed authority over their own planning, nor does a doctor’s assistant have proper insight on the patients’ whereabouts. This solution direction doesn’t necessary improve or decrease the process efficiency or quality. That all depends on the exact implementation. As one can understand, these kind of change initiatives lead to many uncertainties because there is no information on the realization in the future situation.
The future setup introduces many uncertainties, and induces many changes to both the primary operations, as the supporting activities. Medical staff is not only accountable for the medical process outcome. Management focuses on resource utilization, throughput time, waiting lists, and foremost quality of service. This responsibility is put on the medical staff too. The many uncertainties and changes introduced by the transfer to a new facility, with new information systems, processes, and an organization structure, puts a lot of stress on medical staff. This situation results in reluctance and a non-cooperative setting. The staff only pays attention to potential problems, dooming the project to fail.
SIMULATION
Simulation models are often used as a method in problem solving and decision making (Sargent 2005). In modeling a real word problem there is always a payoff between many types of validity, amongst others the internal and external validity, the educational validity, and representational validity (Tsjernikova 2009). Educational validity means the level of complexity and to what extend the model still can be understood. Representational validity implies the actual model similarity with the real world. In this research, we initially modeled the situation to-be in a very generic, simple, and well understandable model. This generic model contained a high educational validity. In follow-up versions the level of detail increased, and the scope of the model was reduced. Hence, focus in this simulation project moved from initially the educational validity, to the representational validity. The results from the simulation, either in graphs, figures or movie clips are used during workshop settings to inform the participants about indications about problems, uncertainties, and opportunities.
In this case study, we have chosen to apply a simulation of the future setup using a software tool, because of
• The many uncertainties concerning the functionalities and process and governance structure in the future setup. On many key implementation aspects, a decision still had to be made, leaving the solution space wide open.
• Solution design options were not yet defined. For every functionality, process and governance structure in the future setup, there were multiple solutions. These solutions needed to be analyzed, discussed, evaluated, and compared. • The connectedness of the organizational departments created a complex solution space. As discussed when
describing the hospital complexity, departments are interdependent in terms of staff, patients, shared resources, financial budgets, and space capacity. A proposed change in one department can influence many other areas, similar to non-linear feedback loops. Simple statistical analysis would not do in these situations.
• The applied software simulation platform provided powerful two dimensional and three dimensional visualization. This aided enormously to the better understanding by the participants in the interactive workshops.
The software simulation tool provided various forms of interactive input processing. During the analysis of different solutions setups, very easily could be interacted by for instance altering constraints and variables. We applied COTS simulation software, often used to analyze the appropriateness of compound facilities’ physical layout, to model the movements of patients. This might sound slightly awkward, because on first sight there seems to be no relation between physical patient movements, and process setup. In our case study, we used results in the physical domain as one of our key performance indicators. These indicators are affected by the process efficiency, and effectiveness of both the information systems and the governance structure. In health care these physical indicators are often of major importance. We measured the following:
• Occupation rate,
• Utilization of expensive equipment, • Patient throughput,
• And waiting times (both for patients and medical staff)
These measures are incorporated in the simulation design, analyzed by means of this simulation software, and evaluated in workshop sessions with the process participants. In this case study, the effects of a change in the facilities layout, combined
with supporting information systems, and an adapted organizational structure, are analyzed using the software simulation tool.
Initially, patients, visitors, and medical staff are represented as multi-agent actors on a 2D layout. The agents are modeled according to the believe, desire, intention (BDI) design (Rao et al. 1995), hence simulating an individual behavior. In the final version of the simulation some parts of the simulation were converted to a 3D simulation, increasing the educational validity even more (and of course making it even more appealing for staff members to watch). The design of the simulation environment was done in iterative cycles. We initiated the project with building a very general model, which we demonstrated to the participant in a representative group from the organization. In following cycles we further specified the model in accordance to available design options, results of simulation analyses, and evaluation by the hospitals representatives. Because of the abstraction layer which is required when modeling all the outpatient treatments in all departments, we received much feedback considering details. This feedback made us zoom in on specific problematic sections of the layout and processes. The initial model was aimed at displaying patient flows on the future layout of the entire hospital building, in the next steps we further zoomed in on a high throughput corridor, then to a team of outpatient clinics working together, and finally, only one outpatient clinic. The level of detail achieved in the last simulation was more than sufficient to convince the staff about the model’s internal validity.
USED DATA
Extensive empirical data is available about the output performance measure concerning efficiency and throughput time over the year 2008. Data used in this simulation is:
• The patients’ arrival profiles (per year, day, minute) • The medical staff’s consultation and operation planning • The known process description and current and future workflow • The information flows and accompanying information systems • Time of patient arrival and departure (interview data)
• And the future facilities’ layout.
In this research project 18 outpatient clinics are included, serving over 210,000 patient treatments yearly. About 42 FTE medical specialists work in the hospital clinics, supported by approximately 120 FTE nurses and support staff. The rest of this section is used to explain input data for the simulation.
Patient arrival (system data)
On average, 800 patients visit the outpatient clinics each day (see Figure 1). As can be seen, the number of patients resembles seasonal influences. Very obviously, the repetitive week structure is visible in this data. This weekly deviation from a flat line number of patients was as well incorporated in the workshop session, because it might be the root cause of many problems in volume handling.
Figure 1: Number of patients per day in 2008.
In the simulation model, only a day at the time is modeled. Because of this, the scheduled treatment time for every patient is important to identify. For simulation purposes, it is most interesting to include the most crowded days in this year. In the
simulation data from the ten busiest days are averaged, and used as a ‘load testing’ scenario. The profile scheduled patients on these days is displayed in Figure 2.
Figure 2: Profile of scheduled patient treatments on crowded days
Naturally, patients do not enter the hospital on the exact time their treatment is scheduled, they arrive early. In general, the amount of time patients arrive early differs per treatment, per region, and per population. Because it was very hard to make the right assumption here, over 375 patients were interviewed over the course of one week. Each patient’s arrival time, scheduled time at a specific department, actually realized meeting time at that department, and departure from the hospital were marked. On average patients arrived 15 minute before their scheduled consultation, see Table 1. These 15 minutes imply thus the precautionary time patients take to arrive before their treatment is scheduled. Delay of this treatment by means of doctors being late will also be added. As will be demonstrated further on, this delay is almost as influential as patients being early.
Average (min.) Variance (min) Minimum (min.) Maximum (min.) Patient arrival before scheduled
consultation
15 9 0 94
Table 1: Patients arrive early
Consultation and operation planning
Each outpatient clinic has its own current and future planning for patient treatments. These schedules (as in Figure 3) determine the amount of patient treatments that can be handled during each physician’s session. Often distinct patient types are scheduled at one session in order to acquire a high throughput rate - the sequencing method as is discussed in section on the complexity of health care organizations. In Figure 3, the schedule for the ORL (Otolaryngology) outpatient clinic. Each marked square indicates the capacity for a specialist to treat patients with certain diagnosis.
Figure 3: Outpatient clinic consultation schedule
Process description
Each department and each type of patient is subject to a (most often) standardized process flow. For this research largely all available standardized process flows are mapped, both for the current, as for the future situation. However though, hospital
Proceedings of the Sixteenth Americas Conference on Information Systems, Lima, Peru, August 12-15, 2010. 7 processes are not static, they are ever evolving. Thus these process flows are subject to continuously improving medical, logistical, or financial insights. Using these process flows is very important for getting an accurate view on the situation as it is, but it is very hard to map out the future state in a simulation based on these dynamically changing flows.
Figure 4: Example of process flow ophthalmology, standard patient
Figure 4 shows the process flow for a patient enrolling in a medical eye examination (department ophthalmology). As is shown, the patient’s disease is analyzed in two sessions, involving three different roles. Total treatment time is twenty minutes, but the patient has to wait for at least ten minutes between the two sessions (to optimize the eavesdropping agent’s effect).
The information flows and accompanying information systems
Information is transferred from patient to hospital, from department to department, and from staff member to staff member. To clearly identify these information transfers and the accompanying information systems, in the most detailed analyses the information flows are mapped out. Figure 5 shows for instance the information transfer between different medical staff in a radiology scan for a hernia patient (neurosurgery outpatient clinic).
Figure 4: Example of process flow ophthalmology, standard patient
Figure 4 shows the process flow for a patient enrolling in a medical eye examination (department ophthalmology). As is shown, the patient’s disease is analyzed in two sessions, involving three different roles. Total treatment time is twenty minutes, but the patient has to wait for at least ten minutes between the two sessions (to optimize the eavesdropping agent’s effect).
The information flows and accompanying information systems
Information is transferred from patient to hospital, from department to department, and from staff member to staff member. To clearly identify these information transfers and the accompanying information systems, in the most detailed analyses the information flows are mapped out. Figure 5 shows for instance the information transfer between different medical staff in a radiology scan for a hernia patient (neurosurgery outpatient clinic).
Figure 5: Running a radiology scan for a hernia patient
Rothengatter et al. Simulation as a method to reframe problems
Actual patient arrival
More striking was the discrepancy between the scheduled and actual consultation time, caused by doctors’ delay. Of all meetings, 12 percent started early, 16 percent on time, and 68 percent too late, for more detail see Table 2. This deviation from the original planning has a major impact on the accumulated amount of time patients spend inside of the hospital building. When limitations are put on the available space capacity, this might be a point of attention.
Percentage (%) Avg. deviation (min) Min deviation (min) Max deviation (min)
Consultation starts early 12 6 1 45
Consultation starts on time 16 - - -
Consultation starts late 68 12 1 60
Table 2: Deviation from start time patient consultation
The total amount of time spent between the hospital’s walls can be deducted by adding all the data from the patient interviews. Table 3 provides an overview.
Average (min.) Variance (min.) Minimum (min.) Maximum (min.) Total time spend by patient in the
hospital
57 1 8 304
Table 3: Total time spend by patients
When the data from Table 3 is combined with schedule treatments profile in Figure 2, an indicative image can be drawn of the amount of patients that are in the hospital at any given time. The result is in Figure 6.
Figure 6: Total amount of patient in the hospital building (indicative).
By all means, these figures are pure indicative, averaged, and accumulated for every outpatient department in this hospital. Naturally these figures are the result of some arrival distributions, and summed up totals from all the departments, and can be spit and diversified to make up a more accurate picture for individual departments or staff members. Of course, in the software simulation this has been done.
In Figure 7, a schematic view of the patients’ flow is depicted, from entrance to exit. During the consultation with the medical specialist a feedback loop could occur, sending a patient back to the waiting for a brief while.
Figure 7: Patient flow Layout
In the simulation, the layout of the future building is used as the basic layer. The CAD drawing is imported and the layout of the building is transported to simulation environment. Patient, doctors, nurses, and visitors are represented as software agents can move within the boundaries of the building. On this basic layer extra information concerning decisions points (such as information desks), targets (such as examination rooms), and routes can be added.
SIMULATION RESULTS
As is stated before, the design of this simulation is done in iterative cycles. The first model and analysis was build with a broad scope on a high level of abstraction, following models were more focused on problematic areas. The final model, which results were conveyed to the complete hospital staff, was focused on one specific outpatient clinic; ophthalmology. The results from the analysis of this department are considered to be exemplary for every outpatient clinic in this hospital. Because of the high level of detail and the small scope included in this simulation, the results were easier to understand for a broad audience. Furthermore the high level of detail provided a high level of trustworthiness for the medical staff. An overview of the simulation results in this section only discusses the outcomes of the ophthalmology outpatient clinic.
The results of the simulation were:
• Occupation rate of the waiting (with a limitation on available seats) near to the clinic • Occupation rate of this clinics examination room
• Heat map of the corridor adjacent to the clinic
• Movie clips displaying a glimpse of future situation, both in 2D as in 3D Waiting occupation
Because of the limitation in available space capacity and consequent maximal waiting capacity, the occupation rate of this waiting is of high importance to the medical staff. Each outpatient clinic has a maximum availability of 18 seats per waiting. In the simulation model only patients were modeled, accompanying visitors were left out of treatment process. To make up for these persons, the hospital management determined a ‘1 simulation agent – 1.75 real life person’ allocation. Figure 8 and Figure 9 show the occupation rate of the ophthalmology waiting. Both figures show the amount of simulation agents populating the waiting. Hence to come to real life persons, the numbers should be multiplied by 1.75. As shown in Figure 9, the most occupied period is between 14:15 and 15:15 hours. The maximum amount of agents in that time frame does not exceed 9, which implies nearly 16 persons.
Figure 8: Occupation rate waiting, morning only
Figure 9: Occupation rate waiting, afternoon only
Examination rooms
A second concern is the examination room occupation by patients. This concern is also expressed by the medical staff. When the patient logistics system does not function to the full extend, and the patient throughput from the central waiting area (near the central entrance) to the examination rooms falters, medical specialists are waiting for patients. The medical staff considers this situation to be highly undesirable in this hospital. The results in Figure 10 and Figure 11 are very hard to interpret. Primarily a repetitive scheme is depicted, which makes a constant flow of patients very likely. The number of patients treated by the medical staff was certainly acquired during the simulations. Earlier investigations about the occupation rate of examination rooms in this hospital made it clear that a level over 40% was not encountered in any in outpatient clinic. The results from these figures therefore seem to match reality. At least the patient logistics system’s performance does seem to suffice.
Figure 10: Occupation examination room number 1AA174, morning only
Figure 11: Occupation examination room number 1AA174, afternoon only
Heat map
The heat map displays the maximum actor activity in the ophthalmology outpatient clinic on the future building layout. The more the color diverges from green, to blue, to yellow, to red, the more activity is forecasted on a specific area. Figure 12 displays this one outpatient clinic, inclusive of waiting, administrative desk, and examination rooms. These kinds of images are used during workshops to convey an outlook for the future situation.
Figure 12: Heat map (maximum actor activity) ophthalmology outpatient clinic
Movie clips (2D and 3D)
The overall simulation results were presented both in hard data and graphs, but also in movie clips showing a possible outlook on the future facility. Initially we were quite reluctant to demonstrate this view on the results, because of the lack of context. Movie clips about 5 minutes are hard to use to demonstrate the future process flow; it would take hours to get a proper idea of the future situation. However, we noticed remarkable enthusiasm after a brief try out. Although lacking educational, the representational validity was enough to make these movie clips useful. Figure 13 displays shots from the 3D environment we applied.
Figure 13: 3D results from the outpatient simulation Examination rooms
CONVEYING THE RESULTS
In a series of seven iterative workshops we have invited a heterogeneous group of healthcare management, specialists, doctor assistants and nurses to collaboratively inspect the simulation, put it to the test, and assess its implications. This way of working results in four major contributions:
1. Enhancement and validity of the model.
From the beginning the model is built around our perception of the hospital’s reality. After interviewing and collaborative reflection our model is adjusted and in some case substantially due to the differences between how practitioners deploy their tasks and also because of differences between specialties or departments within the hospital.
2 Increase in acceptance, insights and awareness of the medical staff
The medical staff in the beginning took a rather critical perspective versus the simulation method itself, but also against the implementation and change project plan of the hospitals board itself. Gradually this critical attitude changed. Also due to the exemplary behavior and attendance of top management during the workshops, but especially because of the well founded arguments given by the simulation model and the results. At the end the participants reported to have learned substantially on operational processes, planning and use of resources in the new situation.
3 New social interactions and collaborations are initiated
The reorganization in the hospital requires a lot of medical staff to start new collaborations. Doctor assistants for example in the new situation will be organized over multiple specialties. This has a great impact on their social and working environment. An important side effect that was not originally our intention is the initiation of new social interactions through the collaborative workshops. The simulation model and its outcomes gave the participants a common theme for discussion and a start to initiate their first collaborations. A majority of the medical staff reports this to be of a major contributor the success of the project.
4 Rephrasing the original ‘problem’ into a constructive setting
The original problem; how to cope with limited available space, the consequential changes in processes and workflow, was rephrased in a setting on how to make most effective and efficient use of shared resources and improve interdepartmental alignment. The need for interdepartmental collaboration was one of the shared visions on the future situation.
CONCLUSION
In this paper we explain how a simulation can be used during a complex reorganization and information system implementation process. The combination of the simulation model with iterative participative design workshops resulted in an increase of acceptance by the prior critical medical staff (both nurses as well as specialists); in increase in their knowledge and insights in the typical problems of the new hospital lay-out, the appropriate processes and the necessary information and the required competences and behavior in the new work setting. The simulation especially worked in this situation because it was able to depict a highly complex reality in a simplified “SIMS” (both 2D and 3D) like world with still a high amount of educational and representational validity. This combination played an important role in convincing critical specialists and getting onboard doctor assistants and nurses. Perceived problems were rephrased in a more constructive setting. The initial problem setting (the patient logistical problem), was rephrased into the question how to improve interdepartmental alignment and collaboration, and making more efficient use of resources available.
REFERENCES
Boonstra, J.J. "Dynamics of organizational change and learning. Wiley handbooks in the psychology of management in organizations,") 2004.
Cutler, D.M. "Equality, Efficiency, and Market Fundamentals: The Dynamics of International Medical-Care Reform," Journal of Economic Literature (40:3) 2002, pp 881-906.
Duke, R.E., and Greenblat, C. "Principles and practices of gaming simulation," Beverly Hills: SAGE) 1981.
Enthoven, S.A.C. "Health Plan: The Only Practical Solution to Soaring Health Care Costs," Reading, Mass.: Addison-Wesley, 1980.
Feinstein, A.H., Mann, S., and Corsun, D. "Charting the experiential territory: Clarifying definitions and uses of computer simulation, games, and role play.," Journal of Management Development (21:10) 2002, pp 732-744.
Feldstein, M.S. "ECONOMETRIC STUDIES OF HEALTH ECONOMICS," Frontiers of Quantitative Economics: papers invited for presentation at the Econometric Society Winter Meetings, Toronto, 1972) 1974, p 377.
French, W.L., and Bell, C.H. Organization development: Behavioral science interventions for organization improvement Prentice-Hall Englewood Cliffs, NJ, 1984.
Glouberman, S., and Mintzberg, H. "Managing the care of health and the cure of disease-Part I: Differentiation," Health Care Management Review (26:1) 2001, p 56.
Gredler, M. Designing and evaluating games and simulations: A process approach Kogan Page, 1992.
Hafferty, F.W., and Castellani, B. "The Increasing Complexities of Professionalism," Academic Medicine (85:2) 2010, p 288. Hans, E., Wullink, G., van Houdenhoven, M., and Kazemier, G. "Robust surgery loading," European Journal of Operational
Research (185:3) 2008, pp 1038-1050.
Harper, P.R., and Shahani, A.K. "Modelling for the planning and management of bed capacities in hospitals," Journal of the Operational Research Society (53:1), Jan 2002, pp 11-18.
Harris, J.E. "The internal organization of hospitals: some economic implications," The Bell Journal of Economics (8:2) 1977, pp 467-482.
Malone, T.W., and Smith, S.A. "Modeling the performance of organizational structures," Operations Research (36:3) 1988, pp 421-436.
Newhouse, J.P. "An Iconoclastic View Of Health Cost Containment," Health Affairs, 1993, pp. 152-171.
Propper, C., Sutton, M., Whitnall, C., and Windmeijer, F. "Did ‘Targets and Terror’Reduce Waiting Times in England for Hospital Care?," HEALTH CARE (8:2) 2008, p 5.
Rao, A.S., and Georgeff, M.P. "BDI agents: From theory to practice," San Francisco, CA, 1995, pp. 312-319. Sargent, R.G. "Verification and validation of simulation models," Winter Simulation Conference, 2005, p. 143.
Stiglic, G., and Kokol, P. "Patient and staff scheduling multi-agent system," IEEE 3rd International Conference on Computational Cybernetics, 2005. ICCC 2005, pp. 25-28.
Tsjernikova, II "Developing and investigating validity of a knowledge management game simulation model," University of Twente, 2009.
van Oostrum, J.M., Van Houdenhoven, M., Vrielink, M.M.J., Klein, J., Hans, E.W., Klimek, M., Wullink, G., Steyerberg, E.W., and Kazemier, G. "A simulation model for determining the optimal size of emergency teams on call in the operating room at night," Anesthesia & Analgesia (107:5) 2008, p 1655.
Vandaele, N., and De Boeck, L. "Advanced resource planning," Robotics and Computer-Integrated Manufacturing (19:1-2), Feb-Apr 2003, pp 211-218.
